python (12.9k questions)

javascript (9.2k questions)

reactjs (4.7k questions)

java (4.2k questions)

java (4.2k questions)

c# (3.5k questions)

c# (3.5k questions)

html (3.3k questions)

Questions - opencl

Trying to skip recasting causes a race condition in OpenCL

I'm just going to start with code, then explain. Say I have the following host code: import pyopencl as cl import numpy as np ctx = cl.create_some_context() queue = cl.CommandQueue(ctx) host_arr = n...
test-img

Cmark

python

opencl

atomic

pyopencl

Votes: 0

Answers: 1

Latest Answer

If device does not have the capability of atomic increment on single byte targets, then it may never do it for every 3 bytes after each 4-byte-aligned byte. Alignment condition could stop it working e...
test-img

huseyin tugrul buyukisik

OpenCL: array of arrays of variable lengths

I am trying to process an array of arrays of variables lengths with OpenCL 1.2 in C++. In each instance (workitem?) I want to process one sub array. Below I've tried to treat the array of arrays as a ...
test-img

sunaj

c++

arrays

opencl

Votes: 0

Answers: 1

Latest Answer

Updated code that works. I have no clue about the performance of this method. vector<cl::Platform> platforms; cl::Platform::get(&platforms); _ASSERT(platforms.size() > 0); auto platform =...
test-img

sunaj

How would I use Xeon Phi with OpenCL

So I just got myself a Xeon Phi and now Im wondering how could I use it with OpenCL because most of the tutorial uses OpenMP. If I change target device to the Xeon Phi and run the kernel will it work?...
test-img

pou

c++

opencl

xeon-phi

Votes: 0

Answers: 0

test-img

Retrieving number of opencl kernel programs that end with a certain result

My Goal is to get a number of kernels, where it's calculation satisfies a condition. The condition could be value == 0. I tried to find some built-in semaphore function to accomplish this, but I could...
test-img

The A

opencl

compute-shader

opencl-c

Votes: 0

Answers: 1

Latest Answer

As pmdj suggested: OpenCL provides atomic operations, including an atomic increment for global (and local) buffers: atomic_inc. unsigned int atomic_inc(volatile __global unsigned int *p) Even better, ...
test-img

The A

Posts

Questions

Blogs